Optimizing Heart Disease Prediction with Feature Sensitivity and Risk Reduction Strategies Using Machine Learning Classifiers

Coronary heart disease (CHD) is associated with a loss of coronary artery blood supply to the heart. The risk of coronary cardiac disease is decreased in physical activity in epidemiologic studies. Researches have shown that regular exercise and the risk of heart disease have an inverse correlation. Sedentary people have the chance of CHD about twice as high-intensity workouts. Among persons with and without CHD, low physical activity levels have been reported to be significant predictors of all-cause and cardiovascular mortality. Evaluating a heart disease risk prediction model using machine learning classifier by considering a feature, number of hours of exercise per week and average calories burnt per day reduces the risk of developing heart disease. The optimum exercise level to stop CHD, however, is uncertain. Increased levels of exercise increase the risk of heart disease, if the patient increases in age. Exercise, both decrease and increase the risk of heart disease. Assessment of the risk of heart disease is very crucial based on effect of just exercise. Data Stratification is one solution where for a given age group, exercise decreases the risk of developing the disease. Various classifiers are used to assess the model. SVM with RBF kernel, KNN and Random Forest gives better accuracy compared to other classifiers.